# US 2016/0253408 A1 · Computer-Implemented System And Method For Providing Classification Suggestions

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2016/0253408 A1
- Document ID: US20160253408A1
- Status: Published application
- Kind: A1
- Application number: 15/150,382
- Filed: 2016-05-09
- Published: 2016-09-01
- Earliest claimed: 2009-07-28
- Assignees: FTI Consulting, Inc.
- Applicants: FTI Consulting, Inc.
- Inventors: William C. Knight; Nicholas I. Nussbaum
- Web page: https://patentgrep.com/patent/US20160253408A1

## CPC

- G06N 5/02: Knowledge representation; Symbolic representation
- G06F 16/287: {Visualization; Browsing}
- G06F 16/3322: {using system suggestions (G06F16/3325 takes precedence)}
- G06F 16/334: Query execution (filtering based on additional data G06F16/335)
- G06F 16/35: Clustering; Classification
- G06F 16/353: into predefined classes
- G06F 16/355: Creation or modification of classes or clusters
- G06F 16/358: Browsing; Visualisation therefor
- G06F 16/93: Document management systems
- G06F 16/954: Navigation, e.g. using categorised browsing
- G06N 5/047: Pattern matching networks; Rete networks
- G06N 7/01: Probabilistic graphical models, e.g. probabilistic networks
- G06N 20/00: Machine learning
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)

## Abstract

A computer-implemented system and method for providing classification suggestions is provided. A set of uncoded documents is maintained. One of the uncoded documents is selected and compared with a set of reference documents, each associated with a classification. Those reference documents that are similar to the uncoded document are identified. Relationships between the uncoded document and each reference document are identified by counting a number of similar reference documents associated with each different classification. The classification having a highest count of similar reference documents is selected for the selected uncoded document as a suggestion.

## Claims

### Claim 1

A computer-implemented system for providing classification suggestions, comprising:
a database to maintain a set of uncoded documents;
a selection module to select one of the uncoded documents and to compare the selected uncoded document with a set of reference documents each associated with a classification;
a similarity module to identify those reference documents that are similar to the uncoded document;
an identification module to identify relationships between the uncoded document and each reference document comprising counting a number of similar reference documents associated with each different classification; and
a suggestion module to suggest for the selected uncoded document the classification having a highest count of similar reference documents.

### Claim 2 (depends on 1)

A system according to claim 1, further comprising:
a vector module to generate a score vector for each uncoded document and each reference document, wherein the score vectors each comprise one or more terms occurring in that document and a score for each term; and
a comparison module to determine a similarity value for each reference document and the uncoded document by comparing the score vectors of that reference document to the score vector of the uncoded document

### Claim 3 (depends on 2)

A system according to claim 2, further comprising:
a threshold module to apply a predetermined threshold to the similarity values and to identify those reference documents with similarity values that satisfy the predetermined threshold as the reference documents similar to the uncoded document.

### Claim 4 (depends on 1)

A system according to claim 1, further comprising:
a display to display the similar reference documents with the uncoded documents.

### Claim 5 (depends on 1)

A system according to claim 1, further comprising:
a display to display the classifications with the similar reference documents.

### Claim 6 (depends on 5)

A system according to claim 5, further comprising:
a classification display module to differentiate different types of the classifications via at least one of color, symbol, and shape.

### Claim 7 (depends on 1)

A system according to claim 1, further comprising:
a placement module to add the selected uncoded document with the suggested classification to the set of reference documents.

### Claim 8 (depends on 1)

A system according to claim 1, further comprising:
a display to display the uncoded documents as a list, wherein at least the selected uncoded document is displayed with the suggested classification.

### Claim 9 (depends on 8)

A system according to claim 8, further comprising:
a reference selection module to select a further uncoded document from the set displayed in the list;
a similarity display module to display the reference documents similar to the further selected uncoded document; and
a classification receipt module to receive for the further uncoded document one of the classifications associated with one or more of the reference documents similar to the further selected uncoded document.

### Claim 10 (depends on 1)

A system according to claim 1, further comprising:
a distance determination module to determine a distance between the selected uncoded document and each of the similar reference documents; and
a weighting module to weigh the count of similar reference documents associated with each classification based on the distances of the associated similar reference documents.

### Claim 11

A computer-implemented method for providing classification suggestions, comprising:
maintaining a set of uncoded documents;
selecting one of the uncoded documents and comparing the selected uncoded document with a set of reference documents each associated with a classification;
identifying those reference documents that are similar to the uncoded document;
identifying relationships between the uncoded document and each reference document comprising counting a number of similar reference documents associated with each different classification; and
suggesting for the selected uncoded document the classification having a highest count of similar reference documents.

### Claim 12 (depends on 11)

A method according to claim 11, further comprising:
generating a score vector for each uncoded document and each reference document, wherein the score vectors each comprise one or more terms occurring in that document and a score for each term; and
determining a similarity value for each reference document and the uncoded document by comparing the score vectors of that reference document to the score vector of the uncoded document

### Claim 13 (depends on 12)

A method according to claim 12, further comprising:
applying a predetermined threshold to the similarity values; and
identifying those reference documents with similarity values that satisfy the predetermined threshold as the reference documents similar to the uncoded document.

### Claim 14 (depends on 11)

A method according to claim 11, further comprising:
displaying the similar reference documents with the uncoded documents.

### Claim 15 (depends on 11)

A method according to claim 11, further comprising:
displaying the classifications with the similar reference documents.

### Claim 16 (depends on 15)

A method according to claim 15, further comprising:
differentiating different types of the classifications via at least one of color, symbol, and shape.

### Claim 17 (depends on 11)

A method according to claim 11, further comprising:
adding the selected uncoded document with the suggested classification to the set of reference documents.

### Claim 18 (depends on 11)

A method according to claim 11, further comprising:
displaying the uncoded documents as a list, wherein at least the selected uncoded document is displayed with the suggested classification.

### Claim 19 (depends on 18)

A method according to claim 18, further comprising:
selecting a further uncoded document from the set displayed in the list;
displaying the reference documents similar to the further selected uncoded document; and
receiving for the further selected uncoded document one of the classifications associated with one or more of the reference documents similar to the further selected uncoded document.

### Claim 20 (depends on 11)

A method according to claim 11, further comprising:
determining a distance between the selected uncoded document and each of the similar reference documents; and
weighing the count of similar reference documents associated with each classification based on the distances of the associated similar reference documents.
